|B.S. University of Minnesota||1968|
|M.S. Stanford University||1969|
|Ph.D. Stanford University||1974|
|Post-Doctoral Fellow Stanford University||1975|
|Professor of Chemistry, The Pennsylvania State University||1975-2000|
Arthur F. and Helen P. Koskinas Professor of Biochemistry
and Molecular Pharmacology
Solving the Protein Folding Problem
Determining the mechanism by which the amino acid sequence of a protein directs the rapid and efficient folding of proteins to their native, functional forms remains one of the most challenging problems in molecular biophysics. The development of a folding code that specifies the three-dimensional structure adopted by a given sequence would complement the genetic code and complete the central dogma of molecular biology (DNA " RNA " Amino Acid Sequence" Functional Protein). Over the past decade, it has also become clear that failure to fold correctly and quickly can lead to numerous diseases. The goals of our lab are to obtain a molecular-level understanding of the folding mechanisms of several very common motifs and to use those insights to contribute to the development of a general theory of protein folding. More recently, we have become interested in applying these biophysical insights to the development of therapeutics for misfolding diseases.
Ongoing projects in the lab
The (βα)8-barrel, or TIM barrel motif, is the most commonly observed fold in biology. TIM barrels are found in five of the six classes of enzymes, and about 10% of all proteins with known three-dimensional structures contain at least one TIM barrel domain. The canonical (βα)8-barrel contains about 200-260 amino acids and is composed of eight βα-repeat units comprised of a β-strand and an anti-parallel α-helix connected by a βα-loop. The eight parallel β-strands form a cylindrical barrel that is surrounded by a shell of the α-helices. In all known (βα)8-barrels, the catalytic active residues are located at the C-terminal ends of the β-strands and in the βα-loops, while residues maintaining the stability of the fold are found at the opposite end of the barrel. The similar topology and structure observed for (βα)8-barrel proteins often arises from vastly divergent amino acid sequences. Thus, TIM barrel proteins are an excellent system for assessing the contributions of the topology and the sequence in governing the folding pathway and the stability of the native structure. Although topology dictates the basic features of the folding free energy surface, we have found that large clusters of isoleucine, leucine and valine side chains between the β-barrel core and the helical shell correspond closely to cores of stability. Surprisingly, we have also discovered a class of side chains-main chain hydrogen bonds can make very significant contributions to stability. Topology and sequence are both crucial to the folding and stability of these globular proteins.
Tryptophan aporepressor (TR) (Figure 2) is a highly helical dimeric DNA-binding protein. The folding kinetics of TR are much more complex, involving three parallel pathways, with multiple steps along each pathway. A dimeric fragment corresponding to the first three helices that contains the hydrophobic core and the dimerization domain of the protein has been constructed and characterized. The folding kinetics of the TR core are much simpler than the full-length protein, permitting a direct measurement of the second-order rate constant for dimerization. We have currently extended our studies of oligomeric proteins to the folding of the trimeric, carbonic anhydrase-g, which contains a novel b-helix structure.
CheY –like proteins
Another class of βα-repeat proteins is the CheY-like protein family, consisting of small (~ 125 residues) globular proteins that function as response regulators in bacterial two-component signal transduction pathways. Their five repeating βα units are arranged in an α/β/α sandwich topology: a central 5‑stranded β-sheet with helices 2, 3 and 4 on one surface of the sheet and helices 1 and 5 on the opposing surface. Several members of this class, including CheY, NT-NtrC and Spo0F (shown on the left), have been shown to fold via intermediates that are kinetically off the productive folding pathway. While the overall mechanism of folding for these proteins is similar and modulated by their common topology, we hypothesize that the structures of the folding intermediates and the details of the folding free-energy landscape are modulated by the amino acid sequence of the individual proteins. The sequence-local contacts between α-helices 2, 3 and 4 and the central β-sheet, define a large cluster of isoleucine, leucine and valine side chains in this early intermediate and provide a mechanism for the microsecond collapse to a kinetically-trapped species during refolding.
Protein folding and ALS
Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease characterized by the loss of motor neurons in the brain and spinal cord. Generally ALS occurs later in life – the age of onset is generally between 40-60 years. After diagnosis, degeneration is rapid and the majority of patients die within 3 years. The largest subset of familial ALS (fALS) cases has been linked to mutations at the Cu, Zn superoxide dismutase (SOD1) locus. To date, over 150 mutations have been discovered, covering nearly 50% of the sequence and nearly all elements of secondary structure. ALS-variants of SOD1 often lead to the formation of insoluble aggregates in the motor neurons of afflicted individuals. Nevertheless, the conformations of SOD1 responsible for aggregation are unclear and a common molecular explanation for the formation of these aggregates has not been discovered. Recognizing that replacements of amino acids generally destabilize the protein, our lab seeks to quantitatively resolve the differences in the structure and folding properties of ALS-variants of SOD1 in an effort to determine a working model for aggregation and potential toxicity. It is also our intention to use insights into the folding mechanism of SOD1 to design assays for small molecular inhibitors of aggregation.
Our lab utilizes a combination of spectroscopy and molecular biology to probe the folding properties of our various model systems. Rather than seek model systems amenable to a specific technique, our approach uses a broad array of techniques to obtain the maximum amount of information, as well as to ensure the accuracy and general applicability of the results. Furthermore, not all techniques are applicable to all proteins, due to limitations of size, solubility and stability. The broad expertise available in the lab allows for the study of systems otherwise intractable with an individual technique.
Probes of global structure, such as circular dichroism and small angle x-ray scattering, are utilized to measure the extent of collapse, and the shape of various folded and unfolded states. On the other hand, fluorescence spectroscopy, both intensity and lifetime measurements, are used to probe the local structure of tryptophan or other extrinsic fluorescent molecules. In combination, these methods are employed to measure the stability of the fold, as well as to ascertain whether particular regions of the protein are participating in folding intermediates. More detailed information on local and global collapse reactions are obtained by utilizing Förster resonance energy transfer (FRET) techniques. These methods, which leverage the exquisite distance sensitivity of field interactions between two fluorescent molecules, allow the measurement of distances between two positions in a protein molecule during refolding and unfolding.
To probe the extent of secondary structure formation in the native and intermediate states along a folding pathway we employ hydrogen exchange techniques coupled with either NMR or mass spectrometry. These methods exploit the fact that amide protons engaged in secondary structure formation, either by the formation of hydrogen bonds or by burial in the core of the protein, are less accessible to exchange with protons from the solvent. These techniques also hold the intriguing possibility of determining the regions of SOD1 which are involved in aggregation, by measuring the differences in protection patterns observed for the soluble protein compared to the aggregated protein.